Self-reported and clinical periodontal conditions in a group of Eastern European postpartum women

Periodontitis is a highly prevalent condition leading to a continuous destruction of tooth-supporting tissues. It increases the risk for various systemic diseases and adverse pregnancy outcomes. Therefore, screening for periodontitis is important. Screening measures can range from self-reported symptoms to clinical full-mouth periodontal examination. The hypothesis of our study was that self-reported parameters and clinical definition perform equally well in identifying periodontitis patients. The aim of this study was to develop, validate its internal consistency, and evaluate a self-reported instrument against periodontal clinical evaluation for diagnosis of periodontitis in a group of postpartum women, as well as to describe their periodontal status and the risk factors associated with periodontal disease. A cross-sectional study on postpartum women was conducted in a tertiary university hospital, from April 2018 to March 2019. Sociodemographic and behavioral data, periodontal clinical parameters, and self-reported periodontal perception were collected. A 16-item questionnaire was developed to obtain information about perceived periodontal alterations and oral hygiene habits. The utility of the questionnaire was tested against a periodontal diagnosis based on a full-mouth periodontal examination. The questionnaire was applied in 215 postpartum women aged 29.16±5.54 years (mean age (y) ± standard deviation) having the following periodontal status: 16 individuals without periodontal disease (7.44%), 32 individuals with gingivitis (14.88%), 19 individuals with mild periodontitis (8.84%), 132 individuals with moderate periodontitis (61.39%), and 16 individuals with severe periodontitis (7.44%). A significant association was observed between oral hygiene score, smoking status, and periodontal conditions (p<0.05). A significant association between the self-reported items related to “gum swelling”, “halitosis”, “previous periodontal diagnosis” and “previous periodontal treatment” with clinical periodontitis have been identified (p<0.05). Using self-reported questionnaires for detection of periodontal disease was ineffective in our studied population, since self-reported parameters and clinical definition do not appear to perform equally in identifying periodontitis cases. Clinical periodontal examination remains the gold standard for screening. Periodontitis was frequent in our group and the severity was significantly associated with the oral hygiene score and smoking. These results underline the necessity for periodontal clinical examination during pregnancy.

Introduction facilitate ongoing studies to evaluate associations between periodontitis and other diseases and conditions [12,17].
Sparse data concerning self-reported periodontal problems compared with clinical assessment are available [12,13,15,18,19]. To date, no self-reported screening attempts addressing periodontitis have been developed in Romania. It is believed that pregnant women are more vigilant about their own health in order to carry a healthy baby to term. As a result, they are more willing to identify any health-related problems. The hypothesis of our study was that self-reported parameters and clinical definition perform equally well in identifying periodontitis patients. Based on this assumption, our primary objective was to develop, validate its internal consistency, and evaluate a self-reported periodontitis instrument and compare it to periodontal clinical diagnosis in a group of postpartum women. The description of the periodontal status, along with the risk-related factors of this group based on clinical assessment, was the secondary objective.

Study design and participants
A cross-sectional study of postpartum women who delivered a live infant prior to examination was conducted. The study was performed in a tertiary university hospital, Gynecology Clinic 1, Emergency County Clinical Hospital, Cluj-Napoca, after receiving ethical approval from the Hospital Ethics Committee (957/24.10.2017) and "Iuliu Hațieganu" University of Medicine and Pharmacy, Cluj-Napoca (9/12.01.2018). Written informed consent was obtained from each of the participants before they answered the written questionnaire and underwent the clinical periodontal examination. In obtaining informed consent and conducting the evaluations, the study adhered to principles outlined in the Declaration of Helsinki on experimentation involving human subjects.
All subjects completed the survey within the first 72 hours after delivery for general demographic data and behaviors, self-reported periodontal conditions, and oral hygiene habits. The periodontal examination in hospitalization conditions was carried out within the same day when the questionnaire was filled out. Data were collected from April 2018 to March 2019. Every ninth subject admitted at the Gynecology Clinic 1, Emergency County Clinical Hospital, Cluj-Napoca who met the inclusion criteria in this period was selected to participate in the study. If the subject refused to participate in the study, the next admitted patient was selected.

Inclusion and exclusion criteria
Postpartum women were eligible for the study if they were aged � 18 years, able to read and understand Romanian, and had delivered a live infant in our hospital. The exclusion criteria were: 1) women aged < 18 years 2) any systemic disease that could possibly influence the history of periodontitis (e.g. chronic hypertension, pregestational diabetes, chronic inflammatory diseases, etc.); 3) any medical condition requiring antibiotic prophylaxis for dental treatment or systemic antibiotic treatment within the last 3 months; 4) human immunodeficiency virus (HIV) infection.
(smoker, nonsmoker, or ex-smoker). The economic status was evaluated according to the national criteria for the definition of poverty, based on the monthly income per family member.

Questionnaire-collected information
A 16-item questionnaire (Table 1, S1 Appendix) was used to obtain information about perceived periodontal alterations (9 items) and oral hygiene habits (7 items). The self-reported periodontitis symptoms section of the questionnaire contained 9 questions and was created by 3 native-speaking Romanian periodontologist seniors. As a template for the questionnaire, the periodontists began with a previously validated Center for Disease Control/American Academy of Periodontology (CDC/AAP) 8-item self-report tool [13,17]. Four items were adapted from the questionnaire elaborated by Eke et al. [13] [15,[20][21][22][23]. A 3-grade answer scale (Yes, No, or I am not sure) was available for each question.
A former validated questionnaire was readapted to collect oral hygiene behavioral data [24]. The oral hygiene part of the questionnaire contains information about toothbrushing frequency, the intervals of toothbrush changing, the type and use frequency of interdental auxiliary hygiene aids (e.g., mouthwash, dental floss, and interdental brushes), the previous professionally supervised oral hygiene, and the periodic dental check-ups [24].
Our English version of the 9-item self-report periodontitis questionnaire was translated into Romanian independently and blindly by 3 experienced periodontists as well as by a native-speaking certified translator. The obtained translations were compared and discussed, and the first version of the Romanian self-report questionnaire was finalized. As recommended by Beaton et al, a cultural adaptation through consensus on the meaning of the  [20]. The clarity for non-specialists and the comprehensibility of the questionnaire was further assessed on a group of 20 consecutive patients undergoing dental care at the Periodontology Department of "Iuliu Hațieganu" University, Emergency County Clinical Hospital, Cluj-Napoca. This alternative group was selected due to the low adhesion previously observed in postpartum women with respect to participation in clinical studies [25]. After filling in the questionnaire and discussing with a senior expert some changes in the wording and refining in the presentation of the questions were performed. The final version of the Romanian self-report questionnaire was finalized ( Table 1). The validation of the Romanian version of the questionnaire was performed against a periodontal diagnosis based on a full-mouth periodontal examination considered as the gold standard for predictive validity (CDC/AAP case definition) [26,27].

Periodontal examination and diagnosis
The postpartum women included in the study underwent a full-mouth periodontal examination with standard methodology and equipment. The oral health examination was done in natural light supplemented by a top-of-the-range front flashlight using a dental mirror and a 1-mm marking periodontal probe (UNC-15 periodontal probe, Hu-Friedy, Chicago, IL, United States). Six sites per each tooth were evaluated for probing depth (PD), gingival recession, and clinical attachment loss (CAL) according to standard clinical definitions [26]. All the teeth (excluding wisdom teeth) yielding 168 sites in a fully dentate individual were evaluated. All probing measurements were rounded down to the nearest millimeter. Bleeding on probing (BoP) was also performed, and a full-mouth BoP score (%) was calculated as the total number of sites with gingival bleeding on probing divided by the total number of sites per mouth (6 sites at each tooth), multiplied by 100 [28]. Oral hygiene was rated using the oral hygiene score (OHS). Scraping was performed in 3 sites at each tooth and calculated as a percentage [29]. The periodontal examination took 30 minutes on average. Maternal postpartum periodontal status was defined initially as a 5-level categorical variable (health, gingivitis, mild -, moderate-and severe periodontitis). A 3-level periodontitis variable (severe, moderate, and mild periodontitis) was established according to CDC/AAP definitions based on measures of CAL and PD at interproximal sites [26,27]. Severe periodontitis was defined as having at least 2 sites with � 6 mm of CAL (not on the same tooth) and at least 1 interproximal site with � 5 mm of PD. Moderate periodontitis was defined as 2 or more interproximal sites with � 4 mm of CAL (not on the same tooth) or 2 or more interproximal sites with � 5 mm of PD, also not on the same tooth. Mild periodontitis was defined as at least 2 interproximal sites with � 3 mm of CAL and at least 2 interproximal sites with � 4 mm of PD (not on the same tooth) or 1 site with � 5 mm. Gingivitis and periodontal health completed the 5-level categorical variable. Patients with gingivitis did not meet the preceding definitions but had a BoP score � 10% [30]. The remaining women were considered clinical periodontal healthy (BoP < 10%) [30].
For the aforementioned periodontal conditions, OHS and missing teeth were reported as the median (first to third quartile) and residual roots as minimum to maximum range.

Investigator training
Periodontal measurements and recordings were performed by 4 experienced investigators (I.C.M., A.S., A.C., S.I.) whose measurements were calibrated in the presence of a senior periodontist. All investigators attended 2 training and calibration meetings; they received oral and written instructions on the development of the study, measurement techniques, and data compilation sheets and were given their precise role and responsibilities in the study. To evaluate intra-examiner and inter-examiner reproducibility, 4 subjects not involved in the study but who met the enrollment criteria were evaluated on 2 occasions, 24 hours apart. The intra-class correlation coefficients, used as a measure of intra-examiner and inter-examiner reliability, were 0.95 and 0.94, respectively.

Data analysis
Experimental data were analyzed with the Statistica program version 13 (StatSoft, Tulsa, OK, United States) at a significance level of 5%. The internal consistency of the proposed tool in the identification of periodontitis was tested with Cronbach alpha after splitting the sample into 1) odd-vs. even-numbered items and 2) 30 randomly determined split-half designs with the correlation of the subtotals for 8 and 7 of the scoring items. Excellent internal consistency is present for Cronbach alpha � 0.9, good for values � 0.8 and < 0.90, acceptable for values � 0.7 and < 0.8, and questionable for values � 0.6 and < 0.7 [31].
The number and the associated percentage were used to report the qualitative data. Chisquare test or Fisher exact test was used to test the association in contingency tables. Quantitative data were reported as median and the value of the first to third quartile (provided in round brackets) and compared using the Kruskal-Wallis test at a significance level of 0.01 (adjusted to the number of groups) due to the number of patients in each subgroup (health, gingivitis, mild periodontitis, moderate periodontitis, and respective severe periodontitis). The variability of OHS among groups was plotted using a box and whisker graph, and comparison between the 2 groups was tested with the Mann-Whitney U test. The ability of individual questions to serve as predictors for periodontitis was tested and reported following the guidelines presented by Bolboacă [32].

Reliability analysis
The final version of the questionnaire was applied in a group of 215 postpartum women. Cronbach alpha applied to the history of dental health problems (9 items) and to oral hygiene behavior (6 items, the frequency of visiting the dentist not considered) had a value of 0.729 and a standardized value of 0.722. The estimated reliability of the scale for adding 8 new questions to the survey is 0.80, under the assumption of a similar inter-correlation between new items. A split-half reliability coefficient equal to 0.729 was obtained for the first strategy of internal reliability analysis (even vs. odd item numbers). The average of the split-half reliability coefficient calculated on 30 random half splits was 0.734, showing the extent of equal contribution of the items to the identification of the periodontitis.

Descriptive analysis-General characteristics and periodontal status of the study group
The eligible population was comprised of 244 women, but 29 refused to participate in the study leading to a participation rate of 88%. Two-hundred fifteen postpartum women with a mean age (y) ± SD (29.16 ± 5.54) fulfilled the inclusion criteria, completed the 2-part questionnaire, and received a full-mouth periodontal examination.
Descriptive sociodemographic data in different periodontal conditions are presented in Table 2. A significant association was observed between smoking status (yes, no, or ex-smoker) and periodontal conditions (healthy, gingivitis, or periodontitis) with a significantly higher percentage of smokers with moderate or severe periodontitis (χ 2 test, 10.5; p value = 0.0330).
The differences regarding the oral hygiene score, missing teeth, and residual roots according to the group are shown in Table 3.
Significant differences were also observed regarding the oral hygiene parameter OHS ( Fig  1). The p values associated to the OHS comparisons were as follows: 0.0087 for health vs. gingivitis; 0.0001 for health vs. moderate periodontitis; <0.0001 for health vs. severe periodontitis; 0.0203 for gingivitis vs. moderate periodontitis; 0.0001 for gingivitis vs. severe periodontitis; 0.0156 for mild periodontitis vs. moderate periodontitis; 0.0003 for mild periodontitis vs. severe periodontitis; and 0.001 for moderate vs. severe periodontitis. Descriptive analysis related to self-reported periodontitis signs in relation to clinical diagnosis is presented in Table 4. The values are expressed as number and percentage according to the group (by column).
Self-reported oral hygiene behaviors are provided in Table 5. The values are expressed as number and percentage according to the group (by column).
A few false-positive results were found indicating periodontal disease where there was no evidence of the disease on the clinical examination. On the contrary, many false-negative results became evident in cases where periodontal disease was found on examination, but the survey indicated otherwise.

Discussion
Due to the ineffectiveness of periodontal epidemiological studies to detect periodontitis in the general population, this study aimed to set up a self-report questionnaire as an accessible and cost-effective tool to track periodontitis in large population groups that could be easily surveyed, such as with pregnant women. We chose this group of patients because of their presumably heightened physical awareness and interest in maintaining their health, for themselves and their future baby. The study also assessed the periodontal status and periodontal risk factors of postpartum women based on the examined clinical parameters.
An increased periodontitis frequency of 77.67% in this group of postpartum women was recorded when compared with the mean frequency of about 50% reported by the literature [1]. This finding may be due to the increased new-onset periodontitis induced by pregnancy [8,9]. The proportion of women with severe periodontitis (7.44%) was lower than that in the general population (11%) [33], probably due to the lower mean age of our group.
From the evaluated periodontitis risk factors, smoking was significantly associated with periodontitis severity, which is to be expected considering the vast amount of data showing an increased prevalence, severity, and progression of periodontitis among smokers in comparison with nonsmokers or ex-smokers [34,35]. Moreover, the outcomes of periodontal treatment are negatively influenced by smoking, but quitting smoking improves the clinical results [35]. The profound effect of smoking on periodontal tissues is related to the modifications of the oral microbiome (new bacteria acquisition through tobacco-infecting flora, increased biofilm formation, and tobacco-induced immunosuppression) [36], and local regenerative capabilities (harmful effects on gingival fibroblasts, mesenchymal stem cells, and bone metabolism) [37,38].
As expected, the values of OHS significantly increased with the severity of periodontal involvement. Although it is a disease with a complex pathogenesis, periodontitis is primarily determined by the aberrant biofilm accumulations on tooth surfaces [39]. Periodontal  inflammation is the natural consequence of dental microbial deposits [40] and it is even more increased during pregnancy [41]. The number of missing teeth is considered a surrogate indicator of periodontitis because it can be viewed as an index of lifetime accumulation of poor oral health [42]. No significant association between tooth loss and periodontitis was identified by the present study. Due to the relatively low mean age of our study group, the severity of the periodontal destruction was not as important to induce tooth loss.
The association analysis revealed significant results between self-report item I.2, "Gum swelling"; I.7, "Halitosis"; I.8, "Previous periodontal diagnosis"; and eventually I.9, "Previous periodontal treatment" and periodontitis ( Table 6). Note that the answers for I.8 and I.9 may reflect a survey bias, namely a socially desirable answer. Furthermore, the access of participants to dental care could induce a bias since most of women were urban residents ( Table 2) and dental medicine offices in urban areas are more common as compared to rural settings. The calculated sensitivity for these 4 items ranged between 10.8% to 31.1%, which is far less than the data of previously published self-report questionnaires for screening periodontitis using the same CDC/AAP case definition. They showed a sensitivity ranging from 44.3% to 85% [13][14][15][43][44][45]. The specificity ranged from 83.3% to 97.9%, which is similar to the values varying from 58% to 92.8% reported by the previously mentioned studies. For the other items, no significant association with periodontal diagnosis was found. Based on our findings, this questionnaire could not accurately detect periodontitis cases.
We only referred to studies having the same CDC/AAP periodontitis case definition because changes in case definition affect the diagnostic accuracy of the self-report questionnaire and hamper the comparability of the results [25]. Carra et al, using the CDC/AAP case definition, communicated a self-report questionnaire validity only for severe periodontitis, although it is difficult to be missed either by patient or practitioner [15]. The present study used supplementary items other than those proposed by Eke et al [13] such as "Bleeding gums" (item I.1). "Bleeding gums", especially bleeding on tooth brushing, is usually related to gingivitis, but also to periodontitis and it is highly exacerbated in periodontally affected pregnant women [41]. In our study "Bleeding gums" did not correlate with periodontitis possibly because its intensity was not proportional with the severity of the periodontal involvement. Gingival bleeding while brushing is the most frequently reported symptom, followed by gingival pain, gingival swelling, spontaneous gingival bleeding, and gingival redness [11]. Generally, the question on bleeding gums had low sensitivity with high specificity, that is, its absence is a good predictor of periodontal health [12]. However, one study showed that self-reported gingival bleeding was correlated with gingival bleeding on clinical examination [19]. Questions on bleeding gums need to be more detailed and a larger study is required to identify its validity [12].
Item I.2 "Gum swelling" was conceived in an effort to identify hyperplasic traits of inflammation associated with gingivitis and periodontitis in pregnancy, which would be identifiable by the patients. Our study showed some association of this item with clinical diagnosis (Table 6). Studies from different countries have shown that approximately 30% to 60% of the participants experience one or more self-report symptoms of gingival inflammation during pregnancy [11]. Items I.3 to I.6 were chosen so as to represent statements of some subjective periodontitis symptoms with an important effect on patients. For example, measures of severe diseases, such as tooth mobility and migration, as well as gingival recessions, may be easier for the patients to notice themselves. Subjects who face more disease might be more aware of their periodontal status [17] and, thus, we would have expected "Visible roots" (I.3), "Tooth migration" (I.4), "Tooth mobility" (I.5), and "Tooth loss" (I.6) to be valid measures of clinical periodontal status, especially of severe disease, which was not the case. Questions regarding tooth mobility and tooth migration had high specificity, but with different values of sensitivity [12].
The low sensitivity values recorded by our study for items I.3 to I.6 may be due to the fact that less than half of women with severe and moderate periodontitis reported the previously mentioned symptoms, that could be a characteristic of the investigated sample with a higher percentage of participants with urban residence and thus higher accessibility to dental medical services. It is possible that periodontitis was exacerbated by pregnancy [8,9] causing important loss of attachment and deep pocket formation but without extensive bone loss leading to other obvious symptoms produced by extensive bone loss (e.g., visible roots, mobility). However, the accuracy and variability of responses to the questions are mediated by population factors, such as literacy, awareness, dental care habits, self-awareness of oral health, access to dental care, age, prevalence, and severity of periodontal disease [12,13].
Halitosis is frequently related to periodontal disease, due to improper oral hygiene but most importantly due to the high proportions of subgingival Gram-negative, anaerobic bacteria which produce different volatile compounds as a result of their metabolic activity [39]. In our analysis, some association between item I.7 and clinical diagnosis was identified, probably because halitosis is a major complaint of most patients with periodontal disease (Table 6).
Item I.8, "Previous periodontal diagnosis" and I.9, "Previous periodontal treatment" have been previously reported to have a good validity [12,13], but they did not fulfill the criteria of reliability to be used as a screening tool in our study (Table 6).
Although the association analysis of some of the items (I.2, I.7, I.8, and I.9) with the clinical periodontitis diagnosis (Table 6) was significant, these items are very poor both for case finding and screening. The important number of false-negatives related to items I.8 and I.9 implies that, although the clinical signs were present, postpartum women were incapable of detecting them. The low sensitivity and clinical utility index values prevent the validation of this selfreport instrument. The AAP advises prenatal care visits as an opportunity for health care providers to educate women regarding behaviors and exposures that might affect their pregnancies and to offer proper periodontal examination and treatment for maintaining good oral health [46]. Periodontitis in pregnant women increases the risk of adverse pregnancy outcomes [5,6,47]. The benefits of intensifying oral hygiene measures during pregnancy should be extensively disseminated [11]. However, few obstetrician-gynecologists advise their patients to seek dental care, and many women decline dental services during this period [48]. No associations between oral hygiene habits and periodontitis were calculated.
One of the limitations of the present study is that the adaptation of some items from a questionnaire specifically developed for a population-based surveillance [13] could not be generalized across different groups. Furthermore, in the same context of questionnaire development, another limitation of our study is related to the validation of the proposed instrument, which was limited to internal consistency. However, the self-report questionnaire developed for the present study was not able to track any disease in our group, not even severe forms. However, the question identifying the hyperplasic inflammatory trait reached some level of significance, but it was singularly analyzed which could be considered as a limitation of the study. It was acknowledged that a single, self-reported item might not be accurate enough to discriminate between individuals with periodontal disease and those without it. Models combining both self-report measures and some demographic characteristics should provide more valid results [12,15,18]. Moreover, there is a great need to educate women in order to increase the perception about potential oral health changes during pregnancy as well as to raise the awareness regarding potential adverse pregnancy outcomes.
However, our self-report questionnaire could be further tested in the Romanian general population with a reliable stratification on age, gender and residence groups and a more accurate reflection of all stages of periodontal disease. Our team considers conducting such a study, after the completion of the research protocol and approval by the ethics committee. Furthermore, the utilization of the current definition of periodontal status [49] was also considered.

Conclusions
Unfortunately, the results of the study prevent the validation of the self-report questionnaire as a screening tool for periodontitis in postpartum women. Self-reported parameters and clinical evaluation do not perform equally well in identifying periodontitis cases in this population group. Until the development of other tools, clinical evaluation remains the gold standard in screening for periodontitis. The frequency of periodontitis in our population group is high, and its severity was significantly associated with two preventable factors: oral hygiene score and smoking.